Performer under stage haze, handheld from the pit
CUTROOM

All tools / Extract Audio From Video

Extract Audio From Video.

Extract the audio from a video file, in the format you actually need.

→ Open the extractorFree, no account, and your video is never uploaded

How it works.

Save just the sound, as an MP3, M4A, or WAV.

  1. Step 1

    Drag your video in

    Or click to choose a file. It stays on your device.

  2. Step 2

    Pick a format, or a range

    MP3 at 128, 192, or 320 kbps.

  3. Step 3

    Save the audio

    One MP3 file, straight to your device.

In detail

One screen, four answers. MP3 for anything that has to play everywhere, M4A when your video already carries AAC and the audio can simply be copied across, WAV when nothing may be lost, and a mono Opus file when the destination is a transcription service rather than a pair of ears. The waveform lets you keep one section instead of the whole recording, and the plan line names the size before anything runs.

What it gives you.

  • MP3 at 128 to 320 kbps

    The number people mean when they say quality. 192 kbps is the default because it is clean on speech and music both, and 320 is there when the source deserves it.

  • Lossless WAV

    16-bit PCM, every sample the decoder produced. It runs about 11 MB a minute in stereo, which is the price of losing nothing.

  • M4A, usually as a straight copy

    Most video already carries AAC audio, and M4A is AAC in the container Apple's apps expect. When that is your source the audio is copied, not re-encoded.

  • A file built for transcription

    Mono Opus at 32 kbps turns a 500 MB video into a few megabytes, which is the smallest thing speech recognition can still hear properly.

  • Trim a range first

    Two marks on the waveform and only that section is extracted. The in and out timecodes end up in the filename so you can tell two clips apart.

  • Runs in your browser

    The file never leaves your machine. No upload, no queue, no watermark, and no stranger holding your recording.

Straight answers.

Which audio format should I extract to?

Start from where it is going. A phone, a car, or a podcast host: MP3. An Apple app, or a video whose audio is already AAC: M4A, which is then a copy and not a re-encode. An editor or anything that will process the sound further: WAV. A transcription or subtitle service: the transcription preset, because those want mono at a low bitrate and reject nothing else about it.

Does extracting audio lose quality?

It depends what you ask for. If the video already carries the codec you want, AAC into an M4A or MP3 into an MP3, the packets are copied byte for byte and nothing is lost. WAV is lossless from the moment the track is decoded. MP3 is a re-encode, so it is a second generation, which at 192 kbps or higher is not something you will hear on speech.

MP3, M4A, or WAV: which one?

MP3 if it has to play everywhere, including a car stereo from 2006. M4A if the audio in your video is already AAC, because then it is a copy rather than a re-encode and it is both faster and cleaner. WAV if it is going into an editor or a mastering chain and the file size does not matter.

How do I send audio to a transcription tool?

Pick For transcription. It writes mono Opus at 32 kbps, which is what speech recognition services actually want: one channel, a low bitrate, and no picture at all. An hour of talking comes out around 14 MB instead of several gigabytes, so the upload is a moment rather than an afternoon.

Will it open my MOV, MKV, or WebM?

Yes. MP4, MOV, MKV, WebM, MP3, M4A, WAV, FLAC, and Ogg are all read by the browser directly. AVI and a few other old containers go through a WebAssembly build of ffmpeg that loads on demand, and so does surround audio like AC3 that browsers cannot decode. The tool says which of those is about to happen before it starts.

How big will the audio file be?

The plan line tells you before you commit. MP3 at 192 kbps is about 1.4 MB a minute, so an hour is roughly 86 MB. Mono Opus at 32 kbps is about 240 KB a minute. Stereo WAV at 48 kHz is 11.5 MB a minute, which is why a two-hour lecture in WAV is over a gigabyte and why the transcription preset exists.

Other tools